The most frequently cited historical parallels, Britain and its 19th-century pax Britannica, or 16th-century Spain, the first country to grasp New World prosperity to dominate the Old World, do not really fit modern America. Both were locked in rivalry with other nearly equal European powers: France and (in the British case) Germany.

Washington readers could do worse than go back to a study of the first real exerciser of unipolar power, the Roman Empire. The book to read is Edward Gibbon's classic study, whose first volume was (by chance) published in 1776, the year of the signing of the American declaration of independence. Gibbon's advice immediately looks quite attractive and relevant to today.

He begins with praise for the peaceful character of the Emperor Augustus and of Roman realism and multilateralism: "Inclined to peace by his temper and situation, it was easy for him to discover that Rome, in her present exalted situation, had much less to hope than to fear from the chance of arms; and that, in the prosecution of remote wars, the undertaking became, every day, more difficult, the event more doubtful, the possession more precarious, and less beneficial."

Previously secure countries can quickly drift away from the political grasp of the hegemon, as resentments and anxieties about the unipolarity of the world grow. Gibbon even had words that might help President George W. Bush understand or deal with Gerhard Schröder's would-be independent foreign policy. "The forests and morasses of Germany were filled with a hardy race of barbarians; and though, on the first attack, they seemed to yield to the weight of Roman power, they soon, by a single act of despair, regained their independence, and reminded Augustus of the vicissitude of fortune." This is a good description of the often counter-productive psychology of the need to slap the hegemon. But such revolt alone is not enough to overthrow hegemonic power.

Why did the Roman version of uni- polarity collapse? Gibbon's empire depended at the height of its success on a sort of multiculturalism, which Romans put in terms of the admission of local deities to the quite crowded complex of the Roman imperial pantheon. In the same way, the US has recently gone out of its way to show how eagerly it will embrace a non-threatening version of Islamic (or indeed Hindu or Confucian) values. A too emphatic insistence on any uniquely Roman virtue or divinity would destroy a precarious notion of cultural pluralism. But it was exactly that plurality of a social and economic kind that proved to be a mechanism of disintegration.

Gibbon saw his story of decline and fall in terms of a revolt against Roman universalism driven by a Christian and egalitarian protest against the unequal distribution of property. This was an early version of an anti-globalisation movement, in which inequalities stemming from the character of imperial power touched off protests. "Most of the crimes which disturb the internal peace of society are produced by the restraints which the necessary, but unequal, laws of property have imposed on the appetites of mankind, by confining to a few the possession of those objects that are coveted by many." Maybe Gibbon was also making a contemporary reference of his own, to the weaknesses of the late-18th-century British empire, with its global commercial culture, so signally exposed and attacked by the American Revolution.

This 18th-century historian's picture of empire and its rules breeding resentment is at odds with a deeply reasoned view, drawn from American social science, about the way in which the modern world can be made peaceful. The argument of the September 2002 US national security strategy initially appears compelling. Because the US can so clearly outspend and defeat (pre-emptively if necessary) any other power, all other powers have an interest in cutting back military spending and threats to their neighbours. This will probably make them wealthier and, therefore, more democratic and peaceful. The world will in this manner be stabilised by the benign force of Washington - a dream very close to that of Augustus.

Which of the arguments is right: the historian's view of long-term decline and fall, or the social scientist's of permanent peace as a result of rational calculation by rational leaders?

The answer depends as much on the stability of the country at the centre as on the behaviour of the rest of the world. The US, unlike the British empire, is building its rule on a foundation that is potentially quite unstable. The British empire in its 19th-century heyday ran enormous current account surpluses (7 per cent of gross domestic product on the eve of the first world war). For more than 20 years, in the period of its cold war victory and of the conversion of the world to a new consensus about markets, the US has had quite large current account deficits. In 2001, the deficit was 4.2 per cent of GDP.

One way of reading this odd situation - which is popular with many Americans - is that the rest of the world has bought into US stability. The deficits are financed by capital inflows, as the non-American world buys the stock of fast-growing US companies or - when the stock market looks bad - property. Indeed, there appears to be a security premium that the rest of the world pays, in that non-American purchases of US assets show consistently lower returns than US purchases of foreign assets.

But nobody thinks that this kind of inflow can be sustained indefinitely. The inflows of foreign capital could be rapidly reversed on some chance piece of bad news. Such a reversal would involve a collapse of the US stock market, the property market and the dollar. US consumers would no longer be able to binge on cheap goods supplied by the rest of the world. American producers would try to protect their markets; foreign producers would be thrown out of business and no longer see any gains to be realised by peaceful integration in a benign world economy.

The financial reversal would also bring the collapse of the US security policy and of its calculated strategy of world pacification. The cost of US defence spending would look much too high and scaling it down would give a chance to would-be rivals, at least on a regional basis - China, for example.

The American case would then look more like that of Spain (which also ran a current account deficit, financed by the outflow of precious metals from its imperial possessions) than that of 19th-century Britain. And Gibbon's story of decline would begin.

The writer is professor of history at Princeton University and author of The End of Globalization